CN116232419A - A nonlinear satellite channel equalization method and system - Google Patents

Abstract

The invention relates to the technical field of satellite-ground broadband communication, and discloses a nonlinear satellite channel equalization method and a nonlinear satellite channel equalization system, wherein the method comprises the following steps: s1, AD sampling: sampling the transmitting signal by adopting an analog-to-digital conversion circuit, and recovering the analog signal into a digital signal; s2, pre-equalization: canceling nonlinear distortion of the sampled digital signal; s3, demodulation: the pre-equalized digital signal is demodulated. The invention solves the problems of poor performance, large operand, poor working stability and the like in the prior art.

Description

A nonlinear satellite channel equalization method and system

Technical Field

The invention relates to the technical field of satellite-ground broadband communication, and in particular to a nonlinear satellite channel equalization method and system.

Background Art

Satellite-to-ground broadband (bandwidth greater than 100MHz) communication technology is one of the core technologies of satellite applications. On the one hand, due to the long distance between satellite and ground, the system uses high-power amplifiers, which causes signal distortion, mainly amplitude modulation/amplitude modulation (AM/AM) and amplitude modulation/phase modulation (AM/PM). On the other hand, in order to improve transmission efficiency, the system often uses high-order signal modulation systems. High-order modulated signals are more susceptible to channel distortion, which reduces reception performance.

Figure 7 describes a typical satellite-ground communication system. The satellite modulates and transmits the data sequence to be sent to generate an analog signal; after passing through the channel, the signal is distorted due to the influence of nonlinear characteristics such as the group delay and nonlinear amplification of the channel. This distortion has little effect on conventional modulation (BPSK (binary phase shift keying), QPSK (quaternary phase shift monitoring), etc.), and the receiving demodulator can achieve good signal reception; but for high-order modulation signals (8PSK (octal phase shift monitoring), 16QAM (hexadecimal quadrature amplitude modulation), 16APSK (hexadecimal amplitude phase modulation signal), 32APSK (thirty-two binary amplitude phase modulation signal)), the channel nonlinearity will have a greater impact on its demodulation sensitivity, and the higher the modulation order, the more obvious the impact. At this time, it is necessary to use a nonlinear equalization method in the receiving demodulator to solve the characteristic function of the channel and use its inverse function to offset the signal distortion caused by the channel characteristics.

From a practical perspective, nonlinear equalization of satellite channels has the following difficulties: 1. Poor performance; 2. Large amount of computation, which cannot be achieved using conventional electronic devices; 3. Poor working stability.

The equalization measures available in the literature are: 1. They are mainly for linear channels; 2. They use equalization methods based on nonlinear algorithms, which have high constraints on usage scenarios and are not universal; 3. They cannot guarantee working stability. At present, there are no reports on practical technologies related to nonlinearity in the field of satellite-to-ground broadband communications. Most of the research is still in the theoretical simulation stage, or based on computer-based back-end processing implementation.

Summary of the invention

In order to overcome the deficiencies of the prior art, the present invention provides a nonlinear satellite channel equalization method and system to solve the problems of poor performance, large amount of calculation, poor working stability, etc. in the prior art.

The technical solution adopted by the present invention to solve the above problems is:

A nonlinear satellite channel equalization method comprises the following steps:

S1, AD sampling: use analog-to-digital conversion circuit to sample the transmitted signal and restore the analog signal to a digital signal;

S2, pre-equalization: offset the nonlinear distortion of the sampled digital signal;

S3, demodulation: demodulate the digital signal after pre-equalization.

As a preferred technical solution, in step S2, the pre-equalizer uses a multi-phase transversal filtering structure to offset the nonlinear distortion of the received digital signal.

As a preferred technical solution, the pre-equalization includes two working modes, namely a self-calibration mode with human participation and an unmanned working mode: in the self-calibration mode, the transmitter sends a known signal to make all the coefficients of the pre-equalization and post-equalization converge according to the channel characteristics, with the goal of minimizing the mean square error between the received signal and the transmitter signal. When the mean square error of the demodulated signal fluctuates within the set range, the self-calibration mode is completed; in the working mode, the pre-equalization coefficient is fixed, and the input signal is adaptively equalized by adjusting the post-equalization, and the result after equalization is the demodulated output.

As a preferred technical solution, in step S2, in the self-calibration mode, the carrier frequency of the known signal at the transmitting end is an integer multiple of the sampling rate of the analog-to-digital conversion circuit, the symbol rate of the known signal at the transmitting end is an integer multiple of the sampling rate of the analog-to-digital conversion circuit, and the modulation system of the known signal at the transmitting end is high-order modulation.

As a preferred technical solution, in step S2, the pre-equalizer compares the distance between the received signal and the known signal at the transmitting end to obtain a nonlinear equalization error, and uses the error to adjust the coefficient of the equalization filter.

As a preferred technical solution, the method further comprises the following steps:

S4, post-equalization: a multi-phase transverse filtering structure is used to perform linear equalization on the linear part of the demodulated data.

As a preferred technical solution, the self-calibration mode includes the following steps:

Z1, perform the following nonlinear combination on the M consecutive sampling points:

为M个输入信号的3阶非线性组合向量，

长度为L，i、j、t为参与进行非线性运算的多个输入信号值，k为非线性组合向量的标号，idx为时间顺序上的运算时刻，M为非线性均衡器长度，M为奇数，in,

is the third-order nonlinear combination vector of M input signals,

The length is L, i, j, t are multiple input signal values involved in the nonlinear operation, k is the label of the nonlinear combination vector, idx is the operation time in time sequence, M is the length of the nonlinear equalizer, M is an odd number,

与L长度的横向滤波器系数点乘，获得非线性组合结果：Z2,

Multiply by the L-length transverse filter coefficients to obtain the nonlinear combination result:

Among them, ^zidx is the output of the third-order nonlinear combination, _culine,k is the transverse filter coefficient sequence, and the initial value of the transverse filter coefficient sequence is as follows:

Z3, take 2N consecutive nonlinear combination result data z ^idx and divide them into two groups according to odd and even numbers:

为非线性组合输出的偶数序列，

为非线性组合输出的第偶数个值，

为非线性组合输出的奇数序列，

为非线性组合输出的第奇数个值；in,

is the even sequence output by the nonlinear combination,

is the even-numbered value of the nonlinear combination output,

is the odd sequence output by the nonlinear combination,

is the odd value of the nonlinear combination output;

Z4, respectively, with initialization coefficients

Linear filtering is performed on z ^idx to obtain:

即为接收到的调制信号；

is the received modulated signal;

为接收到的调制信号的实部，

为接收到的调制信号的虚部；Wherein, c _rvrline,k is the equalization coefficient obtained by extracting the influence of the real part of the nonlinear equalization signal on the real part of the input signal, c _iviline,k is the equalization coefficient obtained by extracting the influence of the imaginary part of the nonlinear equalization signal on the imaginary part of the input signal,

is the real part of the received modulated signal,

is the imaginary part of the received modulated signal;

Z5, compares the error between the received modulated signal and the original signal:

Wherein, err _real is the error between the real part of the equalized output signal and the real part of the transmitted signal, txSym _rea1 is the real part of the transmitted signal, err _imag is the error between the imaginary part of the equalized output signal and the imaginary part of the transmitted signal, and txSym _imag is the imaginary part of the transmitted signal;

Z6, update the filter coefficients of the transverse filter structure of the pre-equalization and post-equalization, and the nonlinear transverse filter coefficient update method is:

为第n次更新的均衡器系数，

为第n-1次更新的均衡器系数，μ_uline为前置均衡更新步进，y^idx1+j为从idx1+j开始的连续L个非线性组合的数据，y^{idx2 +j}为从idx2+j开始的连续L个非线性组合的数据；Where n is the number of updates,

is the equalizer coefficient updated for the nth time,

is the equalizer coefficient updated for the n-1th time, μ _uline is the pre-equalization update step, y ^idx1+j is the data of L consecutive nonlinear combinations starting from idx1+j, and y ^{idx2 +j} is the data of L consecutive nonlinear combinations starting from idx2+j;

The linear transversal filter coefficient update method is:

表示第n次更新的均衡器系数，

表示第n-1次更新的均衡器系数，μ_line表示后置均衡更新步进；in,

represents the equalizer coefficient updated for the nth time,

represents the equalizer coefficient updated for the n-1th time, and μ _line represents the post-equalization update step;

切换至工作模式，接入实际信号。Z7, judge whether the pre-nonlinear equalization coefficient has stabilized. The judgment basis is: if the residual error fluctuation is small, and the distance between the points of the QAM constellation diagram is constant; then it is considered that the self-calibration mode is completed, and the nonlinear transverse filter coefficient is recorded.

Switch to working mode and access the actual signal.

As a preferred technical solution, the working mode includes the following steps:

G1, combine the input sampling signals for M consecutive sampling points according to the combination method and order during self-calibration:

表示M个实际输入信号RX_idx的3阶非线性组合向量，RX_idx-i、Rx_idx-j、Rx_idx-t表示参与非线性运算的多个实际输入信号；in,

represents a third-order nonlinear combination vector of M actual input signals RX _idx , RX _idx-i , Rx _idx-j , and Rx _idx-t represent multiple actual input signals participating in nonlinear operations;

与自校阶段获得的非线性均衡系数序列

点乘，获得非线性均衡结果：G2,

The nonlinear equalization coefficient sequence obtained in the self-calibration stage

Point product to obtain the nonlinear equalization result:

Among them, Rz ^idx is the result after nonlinear equalization.

As a preferred technical solution, post-equalization includes the following steps:

S41, performing timing synchronization and carrier synchronization on the signal after nonlinear processing, and recovering the complex baseband signal rxSig ^idx ;

与横向滤波器系数

点乘，得到滤波结果rxSym^idx：S42, take N consecutive complex baseband signals

and the transverse filter coefficients

Dot multiplication to get the filtering result rxSym ^idx :

提取似然估计与滤波结果的差值：S43, perform maximum likelihood estimation on the filtering result to obtain

Extract the difference between the likelihood estimate and the filtered result:

为滤波结果的最大似然估计；Among them, err ^idx is the error between the maximum likelihood estimate and the actual filtering result,

is the maximum likelihood estimate of the filtering result;

进行更新。S44, use err ^idx to adjust the transverse filter coefficients

to update.

A nonlinear satellite channel equalization system, used to implement the nonlinear satellite channel equalization method, comprises the following modules connected in sequence:

AD sampling module: used to sample the transmitted signal using an analog-to-digital conversion circuit and restore the analog signal to a digital signal;

Pre-equalization module: used to offset the nonlinear distortion of the sampled digital signal;

Demodulation module: used to demodulate the digital signal after pre-equalization;

Post-equalization module: It is used to perform linear equalization processing on the linear part of the demodulated data using a multi-phase transverse filtering structure.

Compared with the prior art, the present invention has the following beneficial effects:

The present invention distributes the solution to the channel problem at the transmitting end and the receiving demodulator, and specifically designs the transmitting end signal, thereby reducing the difficulty of realizing equalization at the receiving demodulator alone; the demodulator equalization method is divided into two parts, which reduces the complexity of realization and improves the stability; the pre-equalization adopts two working modes, and the self-calibration mode can realize human supervision intervention, and the performance is good. In the working mode, the channel parameters are no longer interfered by noise, and the stability is high; the pre-equalization uses the Volterra nonlinear adaptive equalization algorithm to achieve a compromise between performance and complexity; the post-equalization is an equalization method for conventional 100-megahertz-level broadband receiving signals, which takes into account stability, realization difficulty and performance.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG1 is a schematic diagram of a circuit of the present invention;

FIG2 is a schematic diagram of a conventional demodulation circuit;

FIG3 is a schematic diagram of a workflow of the receiver self-calibration phase according to the present invention;

FIG4 is a schematic diagram of a working process of a receiver in the working phase of the present invention;

FIG5 is a comparison diagram of received signal quality without and with the present invention;

FIG6 is a comparison diagram of demodulation quality of receivers without and with the present invention;

FIG. 7 is a diagram showing the working principle of a high-speed wireless data transmission system according to the present invention.

DETAILED DESCRIPTION

The present invention will be further described in detail below in conjunction with embodiments and drawings, but the embodiments of the present invention are not limited thereto.

Example 1

As shown in Figures 1 to 7, the present invention provides a nonlinear method for high-order modulation signals in satellite-to-ground broadband communications, which has a certain degree of universality, greatly improves the demodulation performance, and solves the contradiction between performance and computing resource overhead.

The purpose of the present invention is to provide a nonlinear equalization method for high-order modulation signals in satellite-to-ground broadband communication in view of the shortcomings of the above-mentioned prior art, which has wide applicability, has a good improvement in demodulation performance, and solves the contradiction between equalization performance and computing resource overhead. The method is applied to the receiving demodulator of the satellite-to-ground broadband communication system, is a baseband equalization method based on digital signal processing, and is also a method that combines neural network computing technology with parallel adaptive blind equalization technology.

The above-mentioned purpose of the present invention is achieved by the following measures, and has the following technical features: (1) The solution of the problem is distributed in the transmitter and the receiving demodulator. (2) The satellite channel is estimated by sending a specific signal (specific signal center frequency, symbol rate, modulation method, modulation sequence). (3) The receiving demodulator equalization is divided into two parts, the first part processes the nonlinear part (hereinafter referred to as pre-equalization), and the second part processes the linear part (hereinafter referred to as post-equalization). (4) The pre-equalization is as close to the signal arrival position as possible, and has two working modes, one is a self-calibration mode and the other is a working mode. (5) In the self-calibration mode, the pre-equalization uses the Volterra nonlinear adaptive equalization algorithm to adjust the parameters according to the known signal of the transmitting end. (6) In the working mode, the pre-equalization performs fixed compensation on the input signal according to the adjusted parameters. (7) The post-equalization is an equalization method for conventional 100-megahertz broadband receiving signals.

It is worth noting that: Figure 1 adds two parts (two dotted ellipse boxes: pre-equalization, post-equalization) and these two parts are the core of the present invention; the pre-equalization dotted ellipse box contains a solid line box and a dotted line box, the solid line box is running during the operation of the device, the dotted line box is running before the device works and calculates the working parameters in the solid line box, the stage of the dotted line box operation is called the "self-calibration stage", which requires the "transmitting signal" to send specific data. Figure 2 is a principle block diagram of a receiver. Compared with Figure 1, the timing synchronization and carrier synchronization in Figure 2 are combined into a "demodulation module".

The present invention implements nonlinear equalization (the corresponding digital logic circuit is hereinafter referred to as nonlinear equalization) and linear equalization (the corresponding digital logic circuit is hereinafter referred to as linear equalization) of the received signal by adding two parts of digital logic circuits in the demodulation receiver. The two parts of equalization are coordinated by adopting certain operation steps, thereby achieving higher receiving sensitivity.

The nonlinear equalizer of the present invention is located after the analog-to-digital conversion circuit, has a multi-phase transverse filtering structure, and uses a third-order Volterra structure to offset the nonlinear distortion of the received signal.

The linear equalization of the present invention is located after demodulation, has a multi-phase transverse filtering structure, and adopts a decision feedback linear equalization algorithm to correct the linear interference of the received signal.

The certain operation steps described in the present invention are as follows: before the demodulation receiver of the present invention works, it first receives a specific signal, calculates a set of nonlinear equalization coefficients based on the reception result of the specific signal, and then places the nonlinear equalization coefficients in the nonlinear equalization (called the self-calibration mode); during work, the above coefficients are used to perform nonlinear equalization on the actual signal, and the subsequent linear equalization performs adaptive equalization according to the input working signal (called the working mode).

For the specific signal described in the present invention, the carrier frequency is an integer multiple of the sampling rate of the analog-to-digital conversion circuit of the demodulation receiver, the symbol rate is an integer multiple of the sampling rate of the analog-to-digital conversion circuit of the demodulation receiver, and the modulation system is high-order modulation (the digital information carried by each symbol is greater than two bits).

The present invention calculates a group of nonlinear equalization coefficients, compares the distance between the received signal and the expected value result by nonlinear equalization, obtains the nonlinear equalization error, and uses the error to adjust the equalization filter coefficient.

The subsequent linear equalization described in the present invention adopts a minimum mean square error adaptive blind equalization structure.

The present invention proposes a broadband (bandwidth greater than 100MHz) satellite channel equalization method, which realizes equalization of communication channels and improves receiving sensitivity by adding a specific digital logic circuit to a demodulation receiver and adopting a specific operation process. Before the receiver works, the nonlinear distortion generated by a specific signal passing through a channel and the optimal solution for eliminating the distortion are calculated; when the receiver works, the previous solution is first placed in a dedicated digital logic circuit to realize nonlinear distortion equalization of the input signal, and then a specific digital logic circuit is used to perform linear equalization in real time. The cascade of the two improves the sensitivity of the demodulation receiver. The present invention designs a specific digital modulation signal waveform, and the receiving end can restore the starting sequence without demodulation, compare the difference between the starting sequence and the starting sequence through the channel, and calculate the optimal solution for eliminating distortion. The demodulation receiver adopts a Volterra series digital logic circuit to realize nonlinear equalization of the received signal before demodulation, and then performs conventional demodulation on the signal. After demodulation, a linear equalization digital logic circuit using a minimum mean square error algorithm is used to perform linear equalization on the signal. The present invention combines a specific functional method with a workflow, operates in two dimensions in a coordinated manner, and proposes a nonlinear equalization method for a broadband channel. The method adopts a time domain parallel structure to solve the multiple contradictions of the nonlinear equalization method, which is complex, has poor stability, and has poor effects. The present invention can be applied to broadband transmission signal transmission, remote sensing, and high-speed wireless signal processing.

The present invention relates to the field of satellite-to-ground broadband (bandwidth greater than 100MHz) communication. A broadband signal receiving demodulator uses the method to eliminate the influence of a channel on a high-order modulation signal and improve receiving sensitivity.

Compared with the prior art, the present invention has the following beneficial effects.

The solutions to the channel problem are distributed at the transmitter and the receiving demodulator, and the transmitter signal is specifically designed to reduce the difficulty of achieving equalization at the receiving demodulator alone; the demodulator equalization method is divided into two parts to reduce the implementation complexity and improve stability; the pre-equalization adopts two working modes. In the self-calibration mode, human supervision and intervention can be achieved, and the performance is good. In the working mode, the channel parameters are no longer affected by noise and have high stability; the pre-equalization uses the Volterra nonlinear adaptive equalization algorithm to achieve a compromise between performance and complexity; the post-equalization is an equalization method for conventional 100-megabit broadband receiving signals, which takes into account stability, implementation difficulty and performance.

The method of the invention is suitable for a large-scale logic gate array and is used for a receiving demodulator of a satellite-ground broadband communication system.

It is particularly suitable for satellite-to-ground broadband communication systems that use high-order modulation and are used in receiving demodulators with symbol rates exceeding 100 Mbps.

Example 2

As shown in FIG. 1 to FIG. 7 , as a further optimization of Example 1, based on Example 1, this embodiment also includes the following technical features:

Refer to Figure 1. Compared with the conventional receiving demodulator (Figure 2), the receiving demodulator containing the nonlinear satellite channel equalization method adds two parts: a pre-equalization circuit and a post-equalization circuit. The pre-equalization circuit is located after the analog-to-digital conversion circuit, and performs nonlinear equalization processing on the sampled digital signal to improve the signal quality. The post-equalization circuit is located after the demodulation circuit, and cancels the remaining channel interference, and finally outputs the demodulated signal after equalization.

The receiving demodulator containing the nonlinear satellite channel equalization method has two working modes, namely, a self-calibration mode with human participation and an unmanned working mode. In the self-calibration mode, the transmitter sends a known signal to make all coefficients of the pre-equalizer and the post-equalizer converge according to the channel characteristics. The goal is to minimize the mean square error between the received signal and the transmitter signal. When the mean square error of the demodulated signal fluctuates within a small range, the self-calibration mode is completed. In the working mode, the pre-equalization coefficient is fixed, and the input signal is adaptively equalized by adjusting the post-equalization circuit. The equalized result is the demodulator output.

1. Self-calibration mode:

The transmitter has a known sequence: the signal carrier frequency is one-fourth of the AD sampling rate of the receiving demodulator, the symbol rate is one-eighth of the AD sampling rate, and the modulation mode is fixed to 16QAM. For example, if the AD sampling rate is 4.8GHz, the center frequency of the transmitted signal is 1.2GHz and the symbol rate is 600MHz. The transmitter and the demodulator receiver can have different reference sources, but the relative frequency deviation _fT - _fRfR between the two should not exceed _5.208e ^-7 . This condition can be relaxed to within about 80,000 sampling points, and the frequency difference between the two does not exceed 5.208e ^-7 .

Referring to FIG3 , the following nonlinear combination is performed on M consecutive sampling points (M is an odd number, which is the length of the nonlinear equalizer):

是M个输入信号x_idx+i的3阶非线性组合，共L个。

It is a third-order nonlinear combination of M input signals _xidx+i , a total of L.

与L长度的横向滤波器系数点乘，获得非线性组合结果：

Multiply by the L-length transverse filter coefficients to obtain the nonlinear combination result:

The initial value of the transverse filter coefficient sequence is as follows:

Take 2N consecutive (N is the length of the post-linear equalizer, and 2N means equalizing the real part and the imaginary part respectively) nonlinear combination result data z ^idx , and divide them into two groups according to odd and even numbers.

Initialize the coefficients with

Perform linear filtering on ^zidx .

即为接收到的16QAM调制信号。

That is the received 16QAM modulated signal.

Compare the error between the received signal and the original signal

The coefficients of the two-stage (nonlinear transversal filter and linear transversal filter) filters are updated. The updating method of the coefficients of the nonlinear transversal filter is as follows:

表示第n次更新的均衡器系数，μ_uline表示非线性均衡器更新步进，过大的步进会导致均衡器收敛能力变差甚至不收敛。y^idx1+j，y^idx2+j表示从idx1+j(idx2+j)开始的连续L个非线性组合的数据。

represents the nth updated equalizer coefficient, μ _uline represents the nonlinear equalizer update step, and too large a step will lead to poor convergence of the equalizer or even no convergence. ^{y idx1+j} , y ^idx2+j represent the data of L consecutive nonlinear combinations starting from idx1+j(idx2+j).

Linear transversal filter coefficient update method:

As above, c ⁿ represents the equalizer coefficient updated for the nth time, and μ _line represents the linear equalizer update step. Too large a step will cause the equalizer to have poor convergence ability or even fail to converge.

接收机切换至工作模式，接入实际信号。After about 8,000 symbols of iterative operation, the two sets of transverse filter coefficients will fluctuate around a certain value. At this time, we can manually judge whether the pre-nonlinear equalization coefficients have stabilized. There are two criteria for judgment: the residual error fluctuation is small, and the distance between the points of the QAM constellation diagram is constant. At this time, the self-calibration mode can be considered to be completed, and the nonlinear transverse filter coefficients are recorded.

The receiver switches to working mode and receives the actual signal.

2. The working mode is implemented as follows:

Referring to FIG4 , the input sampling signals are combined according to the combination mode and sequence of the self-calibration for the M consecutive sampling points:

G1, combine the input sampling signals for M consecutive sampling points according to the combination method and order during self-calibration:

表示M个实际输入信号RX_idx的3阶非线性组合向量，Rx_idx-i、Rx_idx-j、Rx_idx-t表示参与非线性运算的多个实际输入信号；in,

represents a third-order nonlinear combination vector of M actual input signals RX _idx , Rx _idx-i , Rx _idx-j , and Rx _idx-t represent multiple actual input signals participating in the nonlinear operation;

与自校阶段获得的非线性均衡系数序列

点乘，获得非线性均衡结果：G2,

The nonlinear equalization coefficient sequence obtained in the self-calibration stage

Point product to obtain the nonlinear equalization result:

Among them, Rz ^idx is the result after nonlinear equalization.

The linear equalizer, its working process is briefly described as follows:

1) Perform timing synchronization and carrier synchronization on the signal after nonlinear processing to restore the complex baseband signal rxSig ^idx .

与横向滤波器系数

点乘，得到滤波结果rxSym^idx。2) Take N consecutive complex baseband signals

and the transverse filter coefficients

Dot product to get the filtering result rxSym ^idx .

提取似然估计与滤波结果的差值3) Perform maximum likelihood estimation on the filtering results and obtain

Extract the difference between the likelihood estimate and the filtered result

4) Use err ^idx to update the transverse filter coefficients. The update algorithm varies depending on the equalization algorithm. See LMS equalization or CMA equalization.

5) rxSym ^idx is the demodulated output result after equalization.

Figures 5 and 6 describe the performance of the method. Figure 5 shows the results of linear and nonlinear equalization after the signal passes through the Saleh channel model; Figure 6 shows the equalization performance of the actual signal after passing through the actual channel. It can be seen that the equalization method described in the present invention can greatly improve the demodulation index.

As described above, the present invention can be preferably implemented.

All features disclosed in all embodiments in this specification, or steps in all methods or processes implicitly disclosed, except for mutually exclusive features and/or steps, can be combined and/or expanded or replaced in any manner.

The above description is only a preferred embodiment of the present invention and does not limit the present invention in any form. According to the technical essence of the present invention, within the spirit and principles of the present invention, any simple modification, equivalent replacement and improvement made to the above embodiment still falls within the protection scope of the technical solution of the present invention.

Claims (10)

1. A nonlinear satellite channel equalization method, characterized in that it comprises the following steps: S1, AD sampling: use analog-to-digital conversion circuit to sample the transmitted signal and restore the analog signal to a digital signal; S2, pre-equalization: offset the nonlinear distortion of the sampled digital signal; S3, demodulation: demodulate the digital signal after pre-equalization. 2. A nonlinear satellite channel equalization method according to claim 1, characterized in that, in step S2, the pre-equalization adopts a multi-phase transverse filtering structure to offset the nonlinear distortion of the received digital signal. 3. A nonlinear satellite channel equalization method according to claim 2, characterized in that the pre-equalization includes two working modes, namely a self-calibration mode with human participation and an unmanned working mode: in the self-calibration mode, the transmitter sends a known signal to make all coefficients of the pre-equalization and post-equalization converge according to the channel characteristics, with the goal of minimizing the mean square error between the received signal and the transmitter signal. When the mean square error of the demodulated signal fluctuates within a set range, the self-calibration mode is completed; in the working mode, the pre-equalization coefficient is fixed, and the input signal is adaptively equalized by adjusting the post-equalization, and the result after equalization is the demodulated output. 4. A nonlinear satellite channel equalization method according to claim 3, characterized in that, in step S2, in the self-calibration mode, the carrier frequency of the known signal at the transmitting end is an integer multiple of the sampling rate of the analog-to-digital conversion circuit, the symbol rate of the known signal at the transmitting end is an integer multiple of the sampling rate of the analog-to-digital conversion circuit, and the modulation system of the known signal at the transmitting end is high-order modulation. 5. A nonlinear satellite channel equalization method according to claim 4, characterized in that, in step S2, the pre-equalization compares the distance between the received signal and the known signal at the transmitting end to obtain a nonlinear equalization error, and uses the error to adjust the equalization filter coefficient. 6. The nonlinear satellite channel equalization method according to claim 5, further comprising the following steps: S4, post-equalization: a multi-phase transverse filtering structure is used to perform linear equalization on the linear part of the demodulated data. 7. A nonlinear satellite channel equalization method according to claim 6, characterized in that the self-calibration mode comprises the following steps: Z1, perform the following nonlinear combination on the M consecutive sampling points:

0≤k≤M-1 i∈[0，M-1] M≤k＜L i∈[0, M-1], j∈[0,i], t∈[0,j]; in,

is the third-order nonlinear combination vector of M input signals,

The length is L, i, j, t are multiple input signal values involved in the nonlinear operation, k is the label of the nonlinear combination vector, idx is the operation time in time sequence, M is the length of the nonlinear equalizer, M is an odd number,

Z2,

Multiply by the L-length transverse filter coefficients to obtain the nonlinear combination result:

Among them, ^zidx is the output of the third-order nonlinear combination, _culine,k is the transverse filter coefficient sequence, and the initial value of the transverse filter coefficient sequence is as follows:

Z3, take 2N consecutive nonlinear combination result data zidx and divide them into two groups according to odd and even numbers:

in,

is the even sequence output by the nonlinear combination,

is the even-numbered value of the nonlinear combination output,

is the odd sequence output by the nonlinear combination,

is the odd value of the nonlinear combination output; Z4, respectively, with initialization coefficients

Linear filtering is performed on z ^idx to obtain:

is the received modulated signal; Wherein, c _rvrline,k is the equalization coefficient obtained by extracting the influence of the real part of the nonlinear equalization signal on the real part of the input signal, c _iviline,k is the equalization coefficient obtained by extracting the influence of the imaginary part of the nonlinear equalization signal on the imaginary part of the input signal,

is the real part of the received modulated signal,

is the imaginary part of the received modulated signal; Z5, compare the error between the received modulated signal and the original signal:

Wherein, err _real is the error between the real part of the equalized output signal and the real part of the transmitted signal, txSym _real is the real part of the transmitted signal, err _imag is the error between the imaginary part of the equalized output signal and the imaginary part of the transmitted signal, and txSym _imag is the imaginary part of the transmitted signal; Z6, update the filter coefficients of the transverse filter structure of the pre-equalization and post-equalization, and the nonlinear transverse filter coefficient update method is:

Where n is the number of updates,

is the equalizer coefficient updated for the nth time,

is the equalizer coefficient updated for the n-1th time, μ _uline is the pre-equalization update step, y ^idx1+j is the data of L consecutive nonlinear combinations starting from idx1+j, and y ^idx2+j is the data of L consecutive nonlinear combinations starting from idx2+j; The linear transversal filter coefficient update method is:

in,

represents the equalizer coefficient updated for the nth time,

represents the equalizer coefficient updated for the n-1th time, and μ _line represents the post-equalization update step; Z7, judge whether the pre-nonlinear equalization coefficient has stabilized. The judgment basis is: if the residual error fluctuation is small, and the distance between the points of the QAM constellation diagram is constant; then it is considered that the self-calibration mode is completed, and the nonlinear transverse filter coefficient is recorded.

Switch to working mode and access the actual signal. 8. A nonlinear satellite channel equalization method according to claim 7, characterized in that the working mode comprises the following steps: G1, combine the input sampling signals for M consecutive sampling points according to the combination method and order during self-calibration:

in,

represents a third-order nonlinear combination vector of M actual input signals Rx _idx , Rx _idx-1 , Rx _idx-j , and Rx _idx-t represent multiple actual input signals participating in the nonlinear operation; G2,

The nonlinear equalization coefficient sequence obtained in the self-calibration stage

Point product to obtain the nonlinear equalization result:

Among them, Rz ^idx is the result after nonlinear equalization. 9. A nonlinear satellite channel equalization method according to any one of claims 5 to 8, characterized in that the post-equalization comprises the following steps: S41, performing timing synchronization and carrier synchronization on the signal after nonlinear processing, and recovering the complex baseband signal rxSig ^idx ; S42, take N consecutive complex baseband signals

and the transverse filter coefficients

Dot multiplication to get the filtering result rxSym ^idx :

S43, perform maximum likelihood estimation on the filtering result to obtain

Extract the difference between the likelihood estimate and the filtered result:

Among them, err ^idx is the error between the maximum likelihood estimate and the actual filtering result,

is the maximum likelihood estimate of the filtering result; S44, use err ^idx to adjust the transverse filter coefficients

to update. 10. A nonlinear satellite channel equalization system, characterized in that it is used to implement a nonlinear satellite channel equalization method according to any one of claims 5 to 9, comprising the following modules connected in sequence: AD sampling module: used to sample the transmitted signal using an analog-to-digital conversion circuit and restore the analog signal to a digital signal; Pre-equalization module: used to offset the nonlinear distortion of the sampled digital signal; Demodulation module: used to demodulate the digital signal after pre-equalization; Post-equalization module: It is used to perform linear equalization processing on the linear part of the demodulated data using a multi-phase transverse filtering structure.

Images